Ethical regulations and publication rules
In a well-functioning democracy, properly conducted opinion polls play an important role. How things are going for a political party or what attitude the Swedish people have on a specific issue is important information for the public to be able to make active choices or to get involved in the public debate. At Demoskop, we are very proud that we were the survey company that came closest to predicting the election outcome in 2018. Since then we have also merged with the survey company Inizio, which was second closest. We are constantly working to improve and ensure the quality of our work. We have our own scientific council for support, but we are active in several industry associations and also collaborate with researchers in Sweden.
Surveys are an excellent way to understand opinions and an exciting way to understand the present. However, it is important to remember that surveys conducted on parts of a population – a sample – are always associated with uncertainty. We at Demoskop help our customers understand what conclusions can be drawn and we also have a responsibility to explain how statistics can be interpreted and used.
Demoskop is a member of several industry associations for leading research companies, both in Sweden and internationally.
The industry association ESOMAR is a global organisation that brings together the world’s leading research companies. The aim is to promote the use of market research and to ensure the quality of implementation. You can read here the ethical code that Demoskop has agreed to follow.
In brief, the ethical code is as follows:
- To set the ethical rules that market researchers must follow with the goal of protecting the respondent
- Raising public confidence in market research by emphasising the rights and protections afforded to the public under these rules
- Participation in a survey must be voluntary for the respondent and researchers must not mislead recipients about the nature of the survey
- The respondent must be guaranteed anonymity, unless otherwise agreed
- The researcher has a duty of confidentiality and must guarantee that all confidential information collected from the respondent is protected
- If children under the age of 14 are to be interviewed, permission must be obtained from an adult
With digitalisation, there are new opportunities to conduct surveys and collect data. Esomar has produced a document with overarching principles that serves as support for research companies on what to think about and how to act.
When collecting, storing and processing data, there will also be a need to clarify what may and may not be done with ‘primary’ data, i.e. the data that we as research companies ourselves collect. Esomar has therefore developed special guidelines for this.
The industry association AAPOR (American Association for Public Opinion Research) focuses on social and opinion polls and arranges lots of interesting seminars and various forms of knowledge sharing. Here you can read their latest update regarding how a survey can be carried out and what the method report should contain.
In Sweden, it is Statistikfrämjandet and its sub-organisation Surveyföreningen that brings together those of us who work with statistics and survey methodology. In order to support survey companies and the media that report surveys, checklists have been produced for what information is needed to be able to ensure the quality of a survey. Here are the checklists for research companies and journalists.
As part of the industry-wide ethical rules that we as survey companies follow, we are obliged to provide information on how the survey was conducted. If the survey company does not want to or cannot provide this information, care must be taken to report the results. However, it is not the case that you must have conducted a certain number of interviews or that the selection must have been made in a certain way. The survey structure depends on what the purpose is and what the target group is. The information that the industry has agreed to include in the delivery of a survey is as follows:
- The name of the survey company and any subcontractors
- Customer’s name (who ordered the survey)
- Purpose of the survey
- The survey’s target group/population that is represented
- Sample size (If the sample size deviates from the planned sample size, the reasons for this must be reported)
- Field work period (when the survey was conducted)
- Selection method and/or recruitment method
- Data collection method
- Response rate/response proportion (if applicable)
- Questionnaire/guide and any display material
- Weighing procedure (how data is processed)
- Remember that surveys are associated with uncertainty.
- Small changes from one measurement to another may be due to chance.
- Shift focus from short-term to long-term changes.
- Report more detailed information about method and sources of error.
- If the survey shows a sensational result that has not been seen in any other survey, take it easy with the dramatic news headlines.
Feel free to contact us if you want to know more.